Unpacking Beauty’s Agentic AI Toolkit

As the wonder trade — and nearly everybody else — come to rely increasingly on AI, they’re starting to embrace the subsequent stage of tech-enhanced client care: AI brokers.

If a big language mannequin is a very smart chatbot, an AI agent is extra like an always-on supercomputer that may observe consumer preferences, provide suggestions for almost any class, from perfume to make-up, and even talk with different AIs to order merchandise or guide providers.

Generative AI requires one human and one machine; agentic AI wants one human however entails a number of machines, defined Wayne Liu, the chief govt of Taiwanese tech agency Perfect Corporation, which launched its Beauty Agent on Monday inside its Youcam app. The firm plans to finally provide its agent as an API, or automated program interface, to manufacturers to allow them to plug-and-play their very own magnificence brokers. (Perfect already affords digital try-on for make-up manufacturers like MAC Cosmetics and La Beauté Louis Vuitton, together with pores and skin and hair evaluation.) Its agent combines these present applied sciences right into a single instrument. To flip it on, “you just start a conversation,” Liu mentioned.

While packing for a visit, you can solicit recommendation from ChatGPT, which might trawl different purchasing lists and guides and provide solutions, or you can ask your magnificence agent, which could reply by asking to take your image. The agent would then use the picture for a pores and skin evaluation, but additionally talk with a big language mannequin, or LLM, for some details about your vacation spot’s local weather to additional personalise its suggestions. Perfect’s digital try-on know-how would additionally permit a consumer to visualise choices for lipsticks or sun shades with out leaving the chat.

Perfect’s new Beauty Agent combines a number of AIs, like picture pores and skin evaluation and LLM-powered suggestions, in a single instrument. (Perfect Corporation)

Agentic AI affords “more like a holistic, autonomous experience” than conversing with an LLM, Liu mentioned.

Widespread adoption remains to be a methods off. Because agentic AI remains to be in its early days, many brokers are solely in a position to execute a handful of actions. Plus, some retailers are already objecting to: Amazon threatened AI agency Perplexity final week, asking it to dam its new agent from searching the e-tailer on a consumer’s behalf.

Consumer manufacturers of every kind are bullish that brokers will create higher buyer experiences and smoother inner workflows. But they’re particularly helpful within the magnificence trade, the place items are more and more bought omnichannel and providers are regularly at hand.

What Is Agentic AI?

AI brokers could be separated into three broad classes: so known as “back-office” brokers that assist with operations and workflow; third-party brokers that use massive language fashions (LLMs) to attract information from completely different sources (and may store on a human’s behalf); and branded brokers, an improve from the AI chatbot “advisors” discovered on many e-commerce websites immediately.

For manufacturers, utilizing agentic AI requires a parallel however separate technique to generative engine optimisation, or GEO. When an individual converses with ChatGPT to offer the 5 best-reviewed foundations launched prior to now decade, what they’re doing is prompting a big language mannequin to reply primarily based on what it could actually scrape from accessible printed information — listicles, magnificence awards, Reddit evaluations. (One examine of 150,000 ChatGPT queries in June 2025 discovered that 40 p.c of solutions pulled from Reddit’s message board.)

An agent, however, would have the option not solely to offer a listing of foundations, however might assist you to buy them or examine native inventory by seamlessly speaking with a back-office agent. They’re “orchestrating in the background with streams of data that are coming from different ports,” mentioned Raakhi Agrawal, a managing director and accomplice at Boston Consulting Group’s tech division BCG.

An agent additionally has a reminiscence, mentioned Liu. “It needs to maintain a persistent identity, so it knows who you are.” It additionally must know when to ask for brand new data primarily based on altering elements like one’s location or routine.

“We don’t assume your skin maintains the same [conditions]because it’s impossible,” Liu mentioned.

What Are Agentic AI’s Applications for Beauty?

Consumer-facing interactions are the obvious functions of agentic AI.

“You can get recommendations on products that are personalised,” Agrawal defined. “If we’re talking about a service, maybe it’s booking through them.”

The hair color firm Madison Reed has experimented with completely different AI instruments since 2014, in line with CEO Amy Everett. This yr, Madison Reed launched Madi, an AI agent, to helps clients guide color appointments in its app.

“I get on Madi and I say I’m in New York City and I need an appointment in Flatiron at 11 o’clock on Saturday; it’s going to tell me I don’t have that, but right away it can also say, I have a spot in Tribeca, in Williamsburg, on 67th and 3rd,” she mentioned. “That person doesn’t have to give us a lot of information.”

Meanwhile, back-office instruments might help in-store employees shortly display stock ranges or talk with different shops and retailers. For manufacturers that already use AI for chatbots or stock planning, utilizing agentic AI can hyperlink these protocols so they impart with each other.

Agrawal sees manufacturers’ largest alternative with agentic AI is in easily integrating their digital and retail experiences.

“You can come, get a skin analysis and a recommendation right away and [then] be able to purchase the product,” Agrawal mentioned. “It would complete that task for you in a way that enhances the experience.”

How Are Beauty Companies Using Agentic AI Right Now?

Today’s brokers can’t do all the pieces; at finest, they’ll do “three things really well, at scale,” mentioned Everett.

For Madi, moreover appointment reserving, that’s helping with buyer care interactions and suggest merchandise primarily based on the model’s rising datasets. The model additionally makes use of back-office instruments to assist stylists anticipate demand for providers weeks prematurely, to allow them to plan staffing and stock.

“That has created more what I would call efficient chair optimisation and better service for the customer,” Everett defined.

The L’Oréal Groupe has partnered with Salesforce AI to include AI brokers into its buyer care interactions, and created an agentic Beauty Genius for the make-up label L’Oréal Paris that may launch on WhatsApp in 2026. Helping shoppers sift by means of suggestions is a precedence for the cosmetics big.

L’Oréal Paris’ Beauty Genius is certainly one of a rising variety of branded brokers. It’ll launch on WhatsApp in 2026. (L’Oreal Paris)

“Beauty is an overwhelming experience… and we could see that consumers get lost and choice paralysis is overwhelming,” mentioned Thomas Vince, chief digital and advertising officer for L’Oréal USA’s Consumer Products Division. The L’Oréal Groupe has additionally invested in an app known as Noli (an acronym for “No One Like I”) to supply extra exact skincare suggestions to customers, fed by the corporate’s a long time of knowledge. The Estée Lauder Companies will quickly unveil its subsequent assistant, a branded perfume suggestion engine, on Jo Malone’s web site.

Demand will seemingly enhance for beauty-specific instruments. While a mannequin like ChatGPT is “very generic,” Liu mentioned Perfect is advertising its AI agent to manufacturers primarily based on its distinctive dataset culled from slightly below 1,000,000 month-to-month lively app customers (who’ve sufficient of a factor for magnificence that they downloaded an app to attempt on lipstick shades).

Not each AI innovation will grow to be commonplace — big-name manufacturers have been experimenting with digital try-on for practically a decade, shoppers nonetheless desire to swatch merchandise in actual life. Agentic AI appear likelier to proliferate, particularly in client care and operations.

These brokers could also be of their infancy, however that doesn’t imply companies ought to wait till they’ve matured; or, as Everett put it, “You’ve got to keep the plane flying while you’re changing out the engine.”

Sign as much as The Business of Beauty publicationyour complimentary, must-read supply for the day’s most essential magnificence and wellness information and evaluation.